Some multiple question choices with more than one correct alternative require all of them to be selected in a final answer. For short and simple examples it is not an issue, since you can just include all of the possibilities in the different alternatives. For longer questions it does become quite cumbersome though, both when creating the question as well as having the students answer them. This article explores a workaround on how to deal with this type of exercise.

Take a look at the following example:

We see that the question has six statements, ranging from a to f. Three of them are correct, a, b and f. In order to get full points, the student needs to give all three of them and none of the incorrect ones.

In detail, the grading works as follows: Each exercise has x correct statements and y incorrect statements. A student selects m + n statements they think are correct. m are actually correct statements (so what the student picked from x) and n are actually incorrect statements (what the student picked from y).
The grade is: (m - n)/(x)
​
Example: A question has six statements, the first 3 are correct.
Student selects 1 and 2:
x = 3
y = 3
m = 2
n = 0
Grade = (2-0)/3 = 2/3

If the student instead selects 1, 2 and 4:
m = 2
n = 1
Grade = (2-1)/3 = 1/3

And so on.

What is possible to realize is that the grade will always be a fraction of the number of correct alternatives. This means that a question with 3 correct statements (x=3) can have 4 possible outcomes (assuming a negative grade is not possible): 100%, 67%, 33% and 0. The way Grasple handles grading, this means that this kind of question has a minimum of 4 different answer rules (one with each different grade). Each extra correct statement in the question adds one extra answer rule. Keep this in mind when adapting your questions to this grading method.

So with all of this in mind it is possible to now actually implement the grading scheme on Grasple. To do so, it is necessary to use the "Substitute" operator in the student answer dependent variables as such:

In short, what is being done here is the following: the student gives their answer as a sum of the different alternatives (a+ b + c for example) and each of the "s" variables is going to substitute one of the alternatives either with +1 or -1 for correct and incorrect alternatives respectively. In the end, "final" calculates what the final grade is.

In the screenshot above, a, b and f are the correct alternatives. So in case of a "a + b + c" answer, the final "s" would be +1 and the final grade 1/3. Each new alternative would add a new "s" parameter, with each correct one adding plus one to the denominator of "final" as well.

The final step is then to use "final" to decide which answer rule to use:

As you can see, there are four answer rules, one for each grade. The way they're chosen is by comparing the "final" parameter with each grade.

Finally, here are a few screenshots of the exercise in practice: